What is ROAS?

How ROAS actually works

The maths is trivial. The reporting is not. Every ROAS figure you look at is really three decisions stacked together: which conversions the platform chose to claim, how far back it was allowed to look, and whether a view counted the same as a click.

Meta defaults to a 7-day click, 1-day view window. Google Ads keeps its own ledger on its own model, and neither can see what the other counted, which is why a Gurugram furniture brand running both will watch their combined reported revenue sail past what Shopify recorded. Nobody is lying. Both counted the same order.

Where teams get ROAS wrong

The most expensive mistake is treating ROAS as a profit number. It is a revenue number. A 4x on an apparel brand running 65 percent gross margin is comfortable, while the same 4x on an electronics reseller working at 12 percent is a slow bleed wearing the costume of a win.

Second, scaling whatever reports highest. High ROAS usually means you found a small pocket of people who had already decided, which is why retargeting your own cart abandoners looks glorious right up until the week you run out of cart abandoners. Then it collapses.

Third, booking revenue before RTO. One in four COD orders coming back turns your 4x into a 3x.

What good looks like in India

There is no universal target. There is a break-even you can work out in ten minutes: take your gross margin percentage, divide 1 by it, and that is the ROAS at which you are standing perfectly still. Forty percent margin means 2.5x.

Then add the leakage nobody puts in the spreadsheet. RTO on COD-heavy categories, returns on apparel, discount codes stacking on each other, and the Shiprocket and Razorpay charges that come off every order before you see a rupee. A Delhi NCR label needing 2.5x on paper often needs 4x in practice.